Accepted for/Published in: JMIR Infodemiology
Date Submitted: Sep 24, 2024
Date Accepted: Feb 13, 2025
Warning: This is an author submission that is not peer-reviewed or edited. Preprints - unless they show as "accepted" - should not be relied on to guide clinical practice or health-related behavior and should not be reported in news media as established information.
Exploring social media posts on lifestyle behaviors: a sentiment and content analysis approach
ABSTRACT
Background:
There has been an increase in the prevalence of non-communicable diseases in Malaysia. This can be prevented and managed through the adoption of healthy lifestyle behaviors, including not smoking, avoiding alcohol consumption, maintaining a balanced diet and being physically active. The growing importance of utilizing social media to deliver information on healthy behaviors has led healthcare professionals (HCPs) to lead these efforts. To ensure effective delivery of information on healthy lifestyle behaviors, HCPs should begin by understanding users’ current opinions about these behaviors and whether the users are receptive to recommended health practices. Nevertheless, there has been limited research conducted in Malaysia that aims to identify the sentiments and content of posts, as well as how well users’ perceptions align with recommended health practices.
Objective:
This study aimed to examine social media posts related to various lifestyle behaviors, by using a combination of sentiment analysis to analyze users’ sentiments and manual content analysis to explore the content of the posts and how well users' perceptions align with recommended health practices.
Methods:
Using keywords based on lifestyle behaviors, posts originating from X and published in Malaysia between November and December 2022 were scraped for sentiment analysis. Posts with positive and negative sentiments were randomly selected for content analysis. A codebook was developed to code the selected posts according to content and alignment of users’ perceptions with recommended health practices.
Results:
A total of 3320 posts were selected for sentiment analysis. Significant associations were observed between sentiment class and lifestyle behaviors (χ2(6)=67.643, P<.001), with positive sentiments higher than negative sentiments for all lifestyle behaviors. Findings from content analysis of 1328 posts have revealed that most of the posts were about users’ narratives (492/1328), general statements (203/1328) and planned actions towards the conduct of their behavior (196/1328). More than half of tobacco-, diet- and activity-related posts were aligned with recommended health practices, whereas most of the alcohol-related posts were not aligned with recommended health practices (63/112).
Conclusions:
As most of the alcohol-related posts did not align with recommended health practices, the findings reflect a need for HCPs to increase their delivery of health information on alcohol consumption. It is also important to ensure the ongoing health promotion of the other three lifestyle behaviors on social media, while continuing to monitor the discussions made by social media users.
Citation
